Hydration of macromolecules. IV. Polypeptide conformation in frozen

Oct 19, 1970 - Bovine albumin. D, urea. FE. 0.445. 0.44 m. Bovine albumin. D,pH 3. FE, T. 0.32. 0.30 l. Hemoglobin. N. FE. 0.415. 0.42 l. Chymotrypsin...
0 downloads 0 Views 385KB Size
516 Table 111. Protein Hydration5 Protein Gelatin Myoglobin Bovine albumin Bovine albumin Bovine albumin Hemoglobin Chymotrypsinogen Lysozyme Lysozyme Ovalbumin TMV coat TMV coat Chymotrypsin Protamine sulfate (-Lysine, 1-glutamate copolymer

Stateb

N N N D, urea D, P H 3 N N

N N N Intact virus Stripped virus. p H 10 N N PH 4 pH 11

Model.

Calcdd

Obsde

FE FE FE FE FE, T FE FE FE FE-B

0.50 0.45 0.445 0.445 0.32 0.415 0.39 0.36 0.335 0.37 0.365 0.365

0.45 0.415 0.40 0.44 0.30 0.42 0.34 0.34 0.34 0.33 0.25 0.361

0.36 0.28

0.33 0.28

FE FE@ FEg FE FE? FE FE

3.0k

2.4k

6.3k

7.Y

Ref 1

f f ,1, m m I

I f f,l

/; I f h h

f f

N, native protein, 0.01 M KC1; D, denatured protein, as shown. c FE, all residues fully exposed; FE. T, residues fully exposed but all carboxylate ions assumed titrated; FE-B, corrected for buried residues. d Residue hydration taken from Table 11. Amino acid analyses from ref 10 and 11. e Results from ref 1 have been revised slightly to yield bovine albumin hydration of0.40,in agreement with later work. f This work, proteins run as described in ref 1, and 2. g R N A is assumed to be unhydrated. Corrected for R N A hydration assumed to be 0.07 g of water per gram of intact virus. W . Bostian and 1. D. Kuntz, unpublished results. 1 Corrected for weight of sulfate as counterion. As moles of water per mole of amino acid. See ref 1. See ref 2. t~

As grams of water per gram of protein.

compositions are available,’O*l 1 few explicit assignments of “protected” and “exposed” residues have been made. l ? Thus, we hake assumed that each amino acid is exposed to the solvent. This should lead to an overestimate of the hydration. The magnitude of this overestimate can be roughly evaluated using lysozyme where the protected residues have been identified.’? Alternatively, one could measure the hydration of completely denatured proteins which should approximate the “fully” exposed state. l 3 We have ignored all contributions from other “structural” features, such as grooves, channels, helical regions, etc. Calculated and experimental (nmr) results are given i n Table 111. Remember that the absolute hydration of both the polypeptides and proteins has been based on the assumption that bovine albumin has a hydration of 0.40 4 0.04 g of water,‘g of protein. The error in the protein measurements is on the order of 10%. The calculated values appear consistently high although within the experimental error. The correction for buried residues is about 10% in the case of lysozyme. This correction should increase as the size of the (globular) protein increases. It is encouraging that both the urea and acid-denatured samples of serum albumin agree with the simple predictions: i.e., the urea exposes the hydrophobic groups which pick up only a small amount of water; acid, by titrating the carboxyl groups, dehydrates the protein. The only striking “structural” effect in Table I11 is a negative one. Tobacco mosaic virus is much less hydrated than the calculations indicate. When the coat protein is dissociated, the hydration increases to the expected value. The comparison of hemoglobin and myoglobin suggests that aggregation of subunits, per se, does not generate a major change in hydration (see also ref 14). We (10) M. 0. Dayhoff, “Atlas of Protein Sequence and Structure,” National Biomedical Research Foundation, Silver Spring, Md., 1969. (11) G. R . Tristam and R. H. Smith, Adoan. Protein Chem., 18, 227 (1963). (12) I . Klotz, Arch. Biochem. Biophys., 138,704 (1970), and references

therein. (13) C. Tanford, Adcan. Protein Chem., 23, 122 (1968). (14) R. Jaenicke and M . A. Lauffer, Biochemistry, 8,3083 (1969).

JournaI of the American Chemical Society

rationalize this finding using the “hydrophobic binding” model for subunit association. If only poorly hydrated hydrophobic regions are involved, one expects small (